Preface
market niches opened by new platforms such as
mobile phones, consoles, tablets, and social media.
Artificial intelligence is a major component contributing to this success, creating the conditions for
more complex virtual environments, realistic nonplayer characters, and engaging experiences.
e core mission of AIIDE is to provide a venue
for industry developers and academic researchers
to exchange ideas, challenges, and solutions. e
scope of the AIIDE conference ranges from the application of artificial intelligence to computer
games all the way to tackling fundamental questions of how intelligent systems can facilitate the
entertainment of humans.
One of the highlights of AIIDE is a series of
keynote talks by prominent industry professionals
and academic researchers. Continuing for a second
year, AIIDE featured a playable experiences track
that showcased fully realized games with significant AI that formed an integral part of player experience.
ese talks and panels frame a peer-reviewed
group of 14 research paper presentations (approximately a 26 percent acceptance rate). In addition,
AIIDE includes a poster and demonstration session
of 15 peer-reviewed papers, one demonstration abstract, and seven playable experiences. Graduate
students had the option of submitting extended abstracts to the ird AIIDE doctoral consortium.
is year’s doctoral consortium selected 5 students
to present their research proposals at the conference and were each assigned a senior researcher as
their mentor.
We would like to thank the members of our Program Committee for their dedicated reviewing service and the Advisory Committee for their continued involvement in guiding the conference. North
Carolina State University served as the gracious
host for the 2014 conference, enabling us to bring
AIIDE to the southeast for the first time. e con-
elcome to the Tenth AAAI Conference on
Artificial Intelligence and Interactive Digital Entertainment (AIIDE-14)! Interactive digital media forms, such as games, are an important part of our society. Beyond entertainment,
interactive and collaborative digital media creates
social bonds between people by providing a shared
reality for people separated in time and space. Interactive media is increasingly used in education and
in ameliorating the cognitive and physical health of
the players. Collaborative media interactions can incentivize crowds of players to solve difficult problems, and researchers now leverage virtual worlds
to serve as a research laboratory for understanding
human behavior.
With the increasing amount of data available on
human behavior within these systems we have an
opportunity to create rich models of various aspects of behavior, such as emotion, skill, and social
preferences. Generative algorithms have the capability to open up novel design spaces of interaction.
is field opens up significant new challenges for
AI systems in their quest to achieve human-level
intelligence. Artificial intelligence in the context of
human entertainment reveals difficult challenges
for researchers and developers that are not found
in other application domains. ese challenges include: creating believable opponents and companions; interactive storytelling; procedural content
generation; player modeling and game analytics;
path planning; and developing tools and representations that empower designers to bring their visions to life.
e computer games industry, the dominant interactive media form in today’s society, has become
a multibillion-dollar commercial enterprise, comparable in size and scope to the film industry.
Game system requirements are an important driver
of hardware and soware innovation in the computer industry, and games have expanded to fill
W
xiii
ference would not have been possible without the
dedicated AAAI Staff, particularly Carol McKenna
Hamilton and Keri Harvey, who helped organize
the conference and are always a pleasure to work
with.
– Ian Horswill, Conference Chair
(Northwestern University)
– Arnav Jhala, Program Chair
(University of California Santa Cruz)
–Organizing Committee
Kevin Dill, Jeff Orkin, Noor Shaker, Mei Si,
Nathan Sturtevant, R. Michael Young
xiv
AIIDE 2014 StarCra Competition
200 games per program pair in round-robin fashion
to increase the statistical significance of the tournament results.To promote the development of strong
AI systems for real-time strategy games, we will
make the source code of tournament entries available to the public at the conclusion of the tournament. is is intended to lower the bar for new entrants to join future competitions, increase the accessibility of AI systems, and promote the exchange
of ideas.
Complete information about the tournament can
be found at starcraaicompetition.com
In 2014, AIIDE will host the Fih Annual StarCra
AI Competition. Participants are given the task of
building the best performing AI system for the popular real-time strategy game StarCra Brood
War (Blizzard Entertainment). e goals of the
competition are to provide a testbed for real-time AI
systems and to promote game AI research by exhibiting AI techniques such as scripting, planning,
optimization, spatial reasoning, and opponent modeling in a fast-paced popular video game.
is year's field includes 19 entries from several
universities and independent StarCra enthusiasts.
ey will play complete games of StarCra Brood
War in a tournament environment similar to those
used by professional gamers. To further replicate
the conditions of professional play, AI systems are
allowed to memorize past encounters with opponents from which to learn and adapt to opponent
strategies. We will utilize 20 lab computers and
tournament scheduling soware to play as many as
– Michael Buro, Competition Cochair
(University of Alberta, Edmonton, Canada)
– David Churchill Competition Cochair
(University of Alberta, Edmonton, Canada)
xv